Abstract

Comprehension and assessments of parameters influencing thermal fatigue damage of high pressure die casting dies

Farid Medjedoub - 10 may 2004

Thermal fatigue is one of the principal life limiting factor of the castings dies in particular high pressure die casting die.

This PhD works deals with an investigation on the effect of the maximum temperature (500, 575, 600, 625, 650 and 685°C, under a constant heat inducted flux) and the heating rate (1.2, 2, 3.7 and 6.5s with respectively 4.81, 3.93, 2.78 et 2.13 MW/m², heat flux density) of the thermal cycle on thermal fatigue behaviour heat checking and cracking of X38CrMoV5 (AISI H11). In situ observations with a contact less long distance microscope (Questar) have revealed that the thermal fatigue cracks remain open during cooling period where the tensile stresses are exercising.

By increasing the heating rate the heat flux density imposed to the cylindrical specimens increases and higher thermal gradient are created. Therefore the thermal fatigue specimens are highly loaded reducing the thermal fatigue life. As the external surface is naturally air cooled, the "mean temperature" of specimen can be changed by modifying the heating rate.

The stress-strain-temperature-time cycles are calculated by Abaqus using thermoelastoplastic constitutive laws (isotropic hardening). The dummy carefully instrumented specimen with several thermocouples (Type K) was used to measure the time-temperature cycles and different locations under various test conditions. These measurements were used in FEM simulations. These curves were used to optimise the transient thermal gradients calculated by Abaqus. A new approach based on normalisation of the thermomechanical characteristics of materials and the thermal cycle is used to estimate the initial heat-flux densities to be imposed on the external surface of the specimen. The thermo elastic stresses are compared with thermal stresses calculated by the normalisation method.

It is observed that the oxidation plays an important role in the formation of the heat-checking (microscopic cracking) and also in macroscopic cracking. It is observed that the heat-checking density presents a sigmodal variation as a function of the number of thermal cycles. The maximum heat-checking density is independent the maximum temperature of thermal cycles, while is very much dependent on the heating rate viz the heat flux density. The heating checking density increases when hating rate increases. It is found that the heat-checking density is changes in as a linear function of the heat flux density. It is observed that beyond a mean stress rate threshold while heating the heat-checking density remains unchanged. A simplified calculation has shown that the difference in heat-checking density can be explained by the difference in the thermomechanical strain of the oxide sc ale and the bulk steel. It is also shown that the heat-checking pattern changes as a function of the local circumferential and longitudinal stresses. A Paris type propagation law can well rationalise all the surface crack propagation rates of macroscopic cracks using Kmax at σ_max.

A fluxmeter using thin thermocouples K (80 µm) was developed in EMAC. This fluxmetre was implemented in critical region of a experimental die mounted on a high pressure die casting machine (610t) for evaluation the transient time-temperature cycle in die under several test conditions (filling rate, solid and liquid lubricant, pressure etc). The time-temperature cycles and the maximum temperature at die surface/Al molten were estimated by inverse method.

Keywords :
thermal fatigue, martensitic steels, hot work tool steels, oxidation, heat-checking, fatigue crack initiation, fatigue crack growth, fatigue life, numerical simulation (ABAQUS™), normalisation method.